Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Gut Online First, published on May 10, 2016 as 10.1136/gutjnl-2016-311818 Oesophagus

ORIGINAL ARTICLE Targeting super-enhancer-associated oncogenes in oesophageal squamous cell carcinoma Yan-Yi Jiang,1 De-Chen Lin,1,2 Anand Mayakonda,1 Masaharu Hazawa,1 Ling-Wen Ding,1 Wen-Wen Chien,1 Liang Xu,1 Ye Chen,1 Jin-Fen Xiao,1 William Senapedis,3 Erkan Baloglu,3 Deepika Kanojia,1 Li Shang,4 Xin Xu,4 Henry Yang,1 Jeffrey W Tyner,5 Ming-Rong Wang,4 H Phillip Koeffler1,6,7

▸ Additional material is ABSTRACT published online only. To view Objectives Oesophageal squamous cell carcinoma Significance of this study please visit the journal online (http://dx.doi.org/10.1136/ (OSCC) is an aggressive malignancy and the major gutjnl-2016-311818). histological subtype of oesophageal cancer. Although recent large-scale genomic analysis has improved the What is already known on this subject? description of the genetic abnormalities of OSCC, few ▸ fi The genomic landscape of oesophageal For numbered af liations see targetable genomic lesions have been identified, and no end of article. squamous cell carcinoma (OSCC) has been molecular therapy is available. This study aims to identify established; however, genetic alterations of Correspondence to druggable candidates in this tumour. actionable targets are infrequent in this Dr Yan-Yi Jiang, Cancer Design High-throughput small-molecule inhibitor malignancy. Science Institute of Singapore, screening was performed to identify potent anti-OSCC ▸ Super-enhancers (SEs) recruit an exceptionally National University of Singapore, NUS, CSI, MD6 14 compounds. Whole-transcriptome sequencing (RNA-Seq) large number of transcription factors/cofactors, Medical drive #13-01, and chromatin immunoprecipitation sequencing (ChIP- and they differ from typical enhancers in size, Singapore 119260, Singapore; Seq) were conducted to decipher the mechanisms of transcription factor density and ability to [email protected] action of CDK7 inhibition in OSCC. A variety of in vitro induce transcription. Dr De-chen Lin, Sun Yat-Sen and in vivo cellular assays were performed to determine ▸ SEs are found to be associated with key Memorial Hospital, Sun Yat- the effects of candidate genes on OSCC malignant lineage-specific master regulators in normal Sen University, Guangzhou phenotypes. somatic cells as well as with a few critical 510120, China; Results The unbiased high-throughput small-molecule [email protected] oncogenes in several types of tumour cells. inhibitor screening led us to discover a highly potent anti- Dr Ming-Rong Wang, Cancer OSCC compound, THZ1, a specific CDK7 inhibitor. RNA- What are the new findings? Institute/Hospital, Peking Union Seq revealed that low-dose THZ1 treatment caused ▸ The SE landscape is established in OSCC cells, Medical College and Chinese and many SE-associated, squamous-specific Academy of Medical Sciences, selective inhibition of a number of oncogenic transcripts. Beijing 100021, China; Notably, further characterisation of the genomic features master regulators and novel oncogenic [email protected] of these THZ1-sensitive transcripts demonstrated that they transcripts are identified. were frequently associated with super-enhancer (SE). ▸ Targeting SE-associated transcription activation Y-YJ, D-CL and AM have by a small-molecule CDK7 inhibitor, THZ1, contributed equally. Moreover, SE analysis alone uncovered many OSCC lineage-specific master regulators. Finally, integrative shows powerful antineoplastic properties Received 8 March 2016 analysis of both THZ1-sensitive and SE-associated against OSCC cells. Revised 11 April 2016 transcripts identified a number of novel OSCC oncogenes, ▸ PAK4 is an SE-associated candidate drug target Accepted 20 April 2016 including PAK4, RUNX1, DNAJB1, SREBF2 and YAP1, in OSCC. with PAK4 being a potential druggable kinase. How might it impact on clinical practice in Conclusions Our integrative approaches led to a the foreseeable future? catalogue of SE-associated master regulators and ▸ fi Through characterising the transcriptional oncogenic transcripts, which may signi cantly promote abnormalities in OSCC, our study suggests that both the understanding of OSCC biology and the targeting transcriptional activation programme development of more innovative therapies. rather than specific genomic lesions might provide better efficacy in the clinical management of solid tumours with highly INTRODUCTION complex genomes, such as OSCC. Oesophageal squamous cell carcinoma (OSCC) is ▸ This work may help establish the potential one of the most common and aggressive GI malig- therapeutic merit of targeting SE-associated nancies.12Due to a lack of understanding of the oncogenic transcription programme through a molecular basis and limited treatment options, the small-molecule CDK7 inhibitor, THZ1, for the prognosis for patients with OSCC has not treatment of patients with OSCC. To cite: Jiang Y-Y, Lin D-C, improved for decades.3 Recently, researchers, Mayakonda A, et al. Gut Published Online First: including ourselves, have determined the genomic 4–8 [please include Day Month landscape of OSCC and identified a number of except those affecting PIK3CA and FGFR1. Year] doi:10.1136/gutjnl- driver events; however, genetic alterations of drug Clearly, alternative molecular approaches are 2016-311818 targets are infrequent in patients with OSCC, needed to further elucidate the pathogenesis of

Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 1 Copyright Article author (or their employer) 2016. Produced by BMJ Publishing Group Ltd (& BSG) under licence. Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus

OSCC for developing more innovative and effective regimens. metastasised into the lungs (around 45 days after initial injec- Transcription factors and cofactors interact with enhancers to tion), they were randomly separated into two groups and control specific gene expression programmes, which are funda- treated with either vehicle or THZ1 (twice daily, 10 mg/kg). At mental to cell biology. Recently, a special group of enhancers, the end of metastasis assays, mice were euthanised and exam- termed super-enhancers (SEs),910have been identified in many ined. Metastasis nodules in lung tissues were fixed in Bouin’’s cell types. SEs recruit an exceptionally large number of tran- solution, embedded in paraffin, cut into 5 mm sections and scription factors/cofactors, and they differ from typical enhan- stained with H&E. All of the animal experiments in this study cers (TEs) in size, transcription factor density and ability to were approved by the Institutional Animal Care and Use induce transcription.11 12 SEs are frequently associated with key Committee (IACUC), National University of Singapore. lineage-specific genes that control cell state and differentiation in normal somatic cells.13 Interestingly, SEs are also found to High-throughput small-molecule inhibitor screening drive the expression of a few critical oncogenes in several types Cell lines were screened for sensitivity against a panel of 104 of tumour cells, such as STAT3,10 MYCN,14 and TAL1.15 small-molecule inhibitors as previously described.17 Briefly, cells However, little is known regarding whether and how SEs drive were exposed to graded concentrations of each drug in 384 well the pathogenesis of OSCC. plates with a seeding density of 250 cells per well in 50 μL final Interestingly, expression of a few SE-associated oncogenes was volume per well. Each drug was tested across an eight-point, reported to be particularly vulnerable to transcriptional inhib- threefold interval dose–response curve (including the no-drug ition, offering a potential cancer targeting approach.12 16 One of control). After 3 days, relative abundance of viable cells was the possible mechanisms is that these SE-associated oncogenes quantified in each well using a tetrazolium-based MTS assay. All are addicted to continuous active transcription, allowing for absorbance values from the MTS assay were normalised to the selective effects before a global blockade of transcription is average of 48 wells per plate containing no drug, and these nor- achieved. malised values were used to fit a third-order polynomial curve In this study, we performed high-throughput small-molecule to each drug dose–response. IC50 values were interpreted from inhibitor screening, and identified THZ1, a newly developed this curve fit to assess the relative sensitivity of each cell line to covalent CDK7 inhibitor,16 as a potent anti-OSCC compound. each drug. Interestingly, we observed that low-dose THZ1 treatment eli- cited selective effects against genes significantly enriched in pro- Chromatin immunoprecipitation sequencing data analysis cesses important in cancer biology. We further characterised the Chromatin immunoprecipitation sequencing (ChIP-Seq) reads SE landscape in OSCC cells and found that THZ1-sensitive were aligned to human reference genome (build GRCh37/hg19) transcripts were significantly more frequently associated with using Bowtie Aligner. ChIP-Seq peaks were identified using SEs. Finally, through integrative analysis of the gene transcrip- MACS (Model-Based Analysis of ChIP-seq) by considering reads tion signature and SE features, we established a functional mapped only once at a given locus. Wiggle files were generated genomic approach to discover novel oncogenes in OSCC. using read pileups for every 50 base pair bins. These wiggle files were normalised in terms of reads per million (rpm) by dividing MATERIALS AND METHODS tag counts in each bin by the total number of reads (in millions, Human cell lines duplicates removed). Wiggle files were converted into bigwig KYSE cell line series were provided by Dr Y Shimada (Kyoto files using wigToBigWig tool (http://hgdownload.cse.ucsc.edu/ University, Japan), and TE-5 and TE-7 cells were provided by admin/exe/) and visualised in Integrative Genomics Viewer Dr Koji Kono (Cancer Science Institute of Singapore, (http://www.broadinstitute.org/igv/home). SEs were identified Singapore). KYSE series cell lines were cultured in RPMI-1640 using ROSE (https://bitbucket.org/youngcomputation/rose). medium; TE-5, TE7 and HEK293T were maintained in Closely spaced peaks (except those within 2 kb of TSS) within a Dulbecco’s modified Eagle medium. All media were supplemen- range of 12.5 kb were merged together, followed by the meas- ted with 10% fetal bovine serum (Invitrogen, San Diego, urement of input and H3K27Ac signals. These merged peaks California, USA), penicillin (100 U/mL) and streptomycin were ranked by H3K27Ac signal and then classified into SEs or (100 mg/mL). TEs. Both SEs and TEs were assigned to the nearest Ensemble genes. The ChIP sequencing files have been deposited into Gene Xenograft assays in NOD-SCID gamma mice Expression Omnibus (GSE76861). Twelve 6-week old, female NOD-SCID gamma (NSG) mice were subcutaneously (s.c.) injected with 1×106 KYSE510 cells Gene set enrichment analysis on their dorsal flanks, with each mouse carrying two explants. Gene set enrichment analysis (GSEA) was performed using After 10–15 days, mice were randomly separated into two GSEA standalone desktop programme. An expression matrix groups and treated with either vehicle or the inhibitor. was created containing expression values at zero and 6 h (upon KPT-9274 was administered orally (150 mg/kg, twice daily, 50 nM THZ1 treatment). All SE-associated genes were used as a 5 days/week), and THZ1 was administered intraperitoneally ‘gene set database’. GSEA was run with parameter ‘Metric for (i.p.) (10 mg/kg, twice daily). The size of the tumours was mea- ranking genes’ set to ‘log2_Ratio_of_classes’ to calculate enrich- sured every 4 days for a total of 4 weeks after treatment. At the ment score for SE-associated genes. end of experiments, mice were euthanised and examined for s.c. xenograft tumour growth. Immunohistochemical (IHC) analysis RESULTS was performed on 5 mm sections of paraffin-embedded s.c. High-throughput small-molecule inhibitor screening tumours. identified small-molecules with potent anti-OSCC properties To identify small-molecule inhibitors with antineoplastic effects Tumour metastasis assay against OSCC cells in an unbiased manner, we first assembled a Twelve 6-week-old female NSG mice were injected with 1×106 focused collection of 104 compounds that had broad targeting KYSE510 cells through the tail vein. Once cancer cells coverage of the kinome.17 18 Four OSCC cell lines were

2 Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus subjected to the high-throughput screen, and IC50 values were CDKs. Notably, all four cell lines were highly sensitive to measured. As a result, 23 inhibitors from different families THZ1, a new covalent CDK7 inhibitor. In addition, OSCC cells showed significant anticancer effect in at least two cell lines were also sensitive to the treatment with flavopiridol and (figure 1A, see online supplementary table S1). Many of the SNS-032, both of which suppressed CDK7 activity among other kinase inhibitors previously shown to have anti-OSCC proper- targets (figure 1B, see online supplementary figure S2). Dose– ties in vitro, including those targeting PI3K/AKT/MTOR response experiments using a panel of 12 OSCC cells showed that 19 20 21 22 23 24 pathway, HSP family or RTK, also demonstrated they were highly sensitive to THZ1 treatment, with IC50 values potent activities here, validating the methodology of our ranging from 21 to 192 nM (figure 1C). We further depleted high-throughput approach (see online supplementary figure S1). CDK7 expression by shRNA-mediated knockdown in TE7 and As CDK inhibitors have shown promising therapeutic merit in KYSE510 and confirmed that CDK7 is essential for both the sur- several other types of tumours but are less well-studied in vival and proliferation of OSCC cells (figure 1D, see online OSCC, we focused on this category of chemicals. To determine supplementary figure S3). As TE7 and KYSE510 cells were among whether specific or pan CDK inhibitors were effective, we the most sensitive ones to THZ1 treatment, we focused our ana- tested a total of seven compounds targeting different lysis on these two cell lines in the following studies.

Figure 1 High-throughput small-molecule screen of four oesophageal squamous cell carcinoma (OSCC) cell lines identified THZ1 with potent anti-OSCC property. (A) Heatmap showed the IC50 values of 104 small-molecule inhibitors in TE7, TE5, KYSE30 and KYSE180 cells. Due to space limitations, only part of the results is shown, and the rest of the results are provided in online supplementary table S1. (B) IC50 values of various CDK inhibitors against OSCC cell lines. (C) Dose–response curves of 12 OSCC cell lines to THZ1 treatment (3 days). Data were represented as mean ±SD of three replicates. (D) TE7 and KYSE510 cells were infected with lentivirus encoding either shRNA against CDK7 or control shRNA (Scramble), and subjected to immunoblotting analysis with indicated antibodies. CC3 denotes Cleaved Caspase-3. (E) Proliferation assay (MTT) showed the effects of THZ1 treatment in OSCC cell lines at indicated time points. Bars represent mean±SD of three experiments performed in triplicate wells (***p<0.001). (F) OSCC cells were treated with either THZ1 or DMSO, and apoptosis was determined by Annexin V–FITC/PI staining. Bars represented mean±SD of three experiments (**p<0.01, ***p<0.001).

Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 3 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus

THZ1 showed powerful antineoplastic properties against increase of apoptosis in the xenografts upon THZ1 administra- OSCC cells tion (figure 2D). Recent studies showed that THZ1 potently suppressed cell pro- We performed additional in vivo experiments to investigate liferation and tumour growth in four different types of malig- further the effect of THZ1 on distal metastasis of OSCC cells. In nancies through inactivation of RNA polymerase II control group, most mice developed visually observable lung meta- (RNAPII)-mediated transcription initiation and elongation, static nodules in 69 days. In contrast, the THZ1-treated mice had including small cell lung cancer,12 neuroblastoma,14 T cell acute no or few tumour nodules in their lungs (figure 2E, F). These lymphoblastic leukaemia16 and triple negative breast cancer.25 results demonstrated that THZ1 possessed very strong antineo- To evaluate the antineoplastic properties of THZ1 against plastic activities against OSCC cells both in vitro and in vivo. OSCC cells, we first performed cell cycle analysis and observed G2/M phase arrest upon THZ1 treatment (see online Global and selective transcription repression by CDK7 supplementary figure S4). THZ1 treatment also resulted in pro- inhibition in OSCC cells found inhibition of cell proliferation and induction of massive We next aimed to understand the mechanisms underlying the apoptosis (figure 1E, F). cytotoxic effects of THZ1 on OSCC cells. CDK7 regulates tran- We subsequently tested the antitumour effects of THZ1 in scriptional processing through recognising and phosphorylating NSG murine models, where each mouse carried two explants the initiation-associated serine 5 (S5) and serine 7 (S7) and formed by KYSE510 cells. The mice received either vehicle or elongation-associated serine 2 (S2) of the RNAPII C-terminal THZ1 twice daily (10 mg/kg). Strikingly, THZ1 completely sup- domain (RNAPII CTD).26 27 Indeed, we observed decreased pressed OSCC tumour growth in vivo (figure 2A–C). phosphorylation S5, S7 and S2 of RNAPII in both TE7 and Importantly, no significant loss of body weight (see online KYSE510 cells in a dose- and time-dependent manner after supplementary figure S5) or other common toxic effects (e.g., THZ1 treatment (figure 3A). We next examined the effect of diarrhoea, rash, etc.; data not shown) were observed. IHC ana- CDK7 inhibition on gene expression profile by performing lysis confirmed the dramatic decrease of cell proliferation and whole-transcriptome sequencing (RNA-Seq) (figure 3B–E, see

Figure 2 THZ1 suppressed the growth of KYSE510 xenografts and inhibited lung metastasis in NOD-SCID gamma (NSG) murine model. (A) Photographs of tumours from both vehicle and THZ1 (twice daily, 10 mg/kg, 24 days) treatment groups. (B) Tumour growth curves of mice treated with either vehicle or THZ1. Data represent mean±SD (****p<0.0001). (C) Weights of tumours from each mice at the endpoint (****p<0.0001). (D) H&E and immunohistochemical (IHC) staining of Ki67, CC3 and Cleaved Caspase-9 (CC9) in tumour tissue sections. Original magnification, ×400. (E) Representative images of lungs after Bouin’’s fixation (left) and H&E staining of tissue sections (right). Original magnification, ×100. Representative tumour lesions are indicated by arrows. (F) Visible lung metastases were counted and graphed (*p<0.05).

4 Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus

Figure 3 THZ1 treatment inhibited RNA polymerase II (RNAPII)-mediated transcription in oesophageal squamous cell carcinoma (OSCC). (A) Immunoblotting analysis of RNAPII C-terminal domain (CTD) phosphorylation in TE7 and KYSE510 cells which were treated with either THZ1 or DMSO at indicated time points (left), with indicated concentrations (right). (B) Heatmap showing expression changes (log2 fold) of all active transcripts upon either DMSO or THZ1 (50 nM) treatment at indicated time points. (C) Box plots of log2 fold changes in global gene expression in TE7 and KYSE510 cells treated with either DMSO or THZ1 (50 nM) at indicated time points. (D) Heatmap and (E) box plots comparing the effect of high-dose (200 nM) and low-dose (50 nM) THZ1 on global mRNA abundance. (F) Enriched gene ontology (GO) functional categories of THZ1-sensitive transcripts.

online supplementary tables S2 and S3). As expected, high-dose which decreased over twofold with low-dose treatment at 6 h) THZ1 (200 nM, i.e., complete inhibition of RNAPII) resulted ‘THZ1-sensitive transcripts’. in global downregulation of steady-state mRNA levels at 6 h Previous studies with other types of THZ1-sensitive cancers (figure 3D, E). Interestingly, we found that low-dose THZ1 found that low-dose THZ1 treatment led to selective inactivation (50 nM, i.e., partial inhibition of RNAPII) elicited downregula- of lineage-specific and oncogenic genes. Based on these finding, tion of a group of transcripts in a gene-selective fashion at early we hypothesised that in the context of OSCC, THZ1-sensitive time points (figure 3B, C). We termed this group (transcripts transcripts might not be random but have important biological

Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 5 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus relevance. Specifically, we speculated that THZ1-sensitive tran- associated with SEs. Interestingly, in other cancer types, scripts might consist of a set of oncogenic transcripts or pathways SE-associated transcripts were found to be particularly sensitive – conferring the sensitivity of OSCC cells to low-dose THZ1 treat- to transcriptional perturbation.9 11 16 Thus, we hypothesised ment. To address this and to analyse comprehensively which sets that these THZ1-sensitive transcripts might be driven by SEs and of genes were particularly sensitive to CDK inhibition, we per- conferred the sensitivity to THZ1 treatment in OSCC. However, formed gene ontology (GO) analysis of THZ1-sensitive tran- SE-associated transcriptional events and their biological rele- scripts. Notably, genes involved in transcription regulation, DNA vance in the context of pathogenesis of OSCC remain unknown. repair and apoptosis regulation were among the most sensitive to CDK7 inhibition in both OSCC cell lines (figure 3F), highlight- Characterisation of SE landscape in OSCC cells ing their crucial roles in OSCC biology. To characterise SE-associated transcripts in OSCC, we per- We next asked whether these THZ1-sensitive transcripts were formed ChIP-Seq using the antibody recognising H3K27ac associated with any genomic features. It had been reported that modifications in OSCC cells which were also profiled by master transcription factors and cofactors which played key roles RNA-Seq. As a result, we annotated 444 and 855 SE-associated in cell identity and malignant phenotypes were frequently transcripts in TE7 and KYSE510 cell lines, respectively (figure

Figure 4 Characterisation of super-enhancer (SE)-associated genes in TE7 and KYSE510 cell lines. (A) Hockey stick plots of oesophageal squamous cell carcinoma (OSCC) cells showing input normalised, rank ordered H3K27ac signals for SE-associated genes. (B) Gene ontology (GO) analysis of enriched biological processes for SE-associated genes in OSCC cells. (C) H3K27ac chromatin immunoprecipitation sequencing (ChIP-Seq) binding profiles of representative SE-associated genes in both cell lines. (D) Left, mRNA expression level of representative transcripts (MIR205HG and IRF6) across various types of human cancer cells; Right, MIR205HG and IRF6 expression in OSCC compared with OA cell lines. Data were retrieved from CCLE project from Broad Institute. (E) Gene set enrichment analysis of THZ1-sensitive and SE-associated transcripts. (F) Box plots showing log2 fold changes for transcripts associated with total pool of all enhancers (ALL), typical enhancers (TEs) and SEs upon THZ1 treatment at 6 h.

6 Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus

4A, see online supplementary table S4). Notably, we readily remained unaltered upon exposure to THZ1 (see online observed that many of these transcripts were lineage-specific supplementary figure S8). transcription factors acting as master regulators of keratinocyte To test their biological functions in OSCC cells, we silenced cell differentiation, such as KLF5,28 TP6329 and IRF630 (figure each one of these transcripts by siRNA-mediated knockdown 4A,C). This was concordant with the fact that both of these cell and performed cell proliferation assay in these two representa- lines are squamous cell type. Furthermore, a number of well- tive cell lines (figure 5C, see online supplementary table S6). As defined OSCC oncogenes were also associated with SEs, such as a result, we identified four candidate genes, namely, RUNX1, TP63,31 EGFR,32 ANO1,33 SOX2,34 FSCN135 and CTTN36 YAP1, DNAJB1 and PAK4, which contributed to the prolifer- (figure 4A, C, see online supplementary figure S6). To determine ation of the two OSCC cells (figure 5D, E, see online this enrichment more comprehensively, we performed GO ana- supplementary figure S9). We also confirmed that these four lysis. Importantly, squamous-cell unique biological processes, proteins were significantly decreased in both cell lines by THZ1 including epidermis development and keratinocyte differenti- treatment (figure 5F). Worthwhile to note, SREBF2 was asso- ation, were highly enriched in both cell lines (figure 4B). In add- ciated with SEs in KYSE510 but not TE7 cells; and protein and ition, genes involved in cancer-related functions such as cell mRNA levels of this gene in KYSE510 cells were consistently proliferation and apoptosis were also significantly overrepre- highly expressed compared with TE7 cells (figure 6A). sented in SE-associated transcripts. Moreover, we identified Importantly, SREBF2 only promoted the growth of KYSE510 many novel SE-associated transcripts, including both coding and but not TE7 cells (figure 5E, see online supplementary figure non-coding RNAs, whose functions have not been reported in S9), underscoring the ability of our approaches to delineate the the setting of OSCC biology.37 For example, non-coding RNA connections between expression pattern, SE feature and bio- MIR205HG was associated with one of the biggest SEs in both logical relevance. cells (figure 4C). As SEs were reported to drive lineage-specific To characterise further the biological functions of these five expression of key transaction factors in somatic cells, we next novel candidate oncogenes in OSCC, we measured their protein analysed and found that a number of SE-associated transcripts, levels in a panel of 12 OSCC cell lines and identified additional such as MIR205HG, IRF6, TP63 and SOX2, showed a lineage- cell lines with high expression of each gene (figure 6A). We next specific expression pattern, with highest expression levels in chose additional representative cell lines which expressed candi- squamous cell cancers, including OSCC and head/neck squa- date genes at high levels for functional assays. Importantly, mous carcinoma (figure 4D, see online supplementary figure MTT and colony formation analysis showed that each of these S7A, B). To confirm further this expression pattern, we com- five genes was required for cell proliferation in at least three dif- pared OSCC with oesophageal adenocarcinoma (OA) and found ferent OSCC cell lines (figure 6B). that many SE-associated transcripts identified in our OSCC cells We next determined whether these five SE-associated onco- were more highly expressed in the squamous cell type (represen- genes are up-regulated in OSCC specimens compared with non- tative results shown in figure 4D, see online supplementary malignant oesophageal epithelium. Importantly, IHC staining figure S7). Together, these results suggest that in OSCC cells, showed that PAK4, SREBF2 and YAP1 proteins were highly SE-associated transcripts contain lineage-specific master regula- expressed in OSCC tissues but were markedly lower in adjacent tors and oncogenic transcripts; many of which displayed tissues- normal oesophageal epithelium (figure 6C). We also examined specific expression patterns. the expression of RUNX1 and DNAJB1 proteins; however, their antibodies failed to generate specific signals in the experi- Identification of novel SE-associated oncogenes in OSCC ments (data not shown). We next asked whether SE-associated transcripts are dispropor- tionately sensitive to CDK7 inhibition. GSEA for all active genes Identification of PAK4 as an SE-associated candidate drug upon low-dose THZ1 treatment showed that the majority of the target in OSCC genes in the leading edge were SE-associated transcripts, and The p21-activated kinases (PAKs) belong to Ser/Thr protein they were most sensitive to THZ1 treatment (figure 4E). kinase family, which contains six members in humans (PAK1–6). Moreover, the abundance of SE-associated transcripts were Overexpression of PAK4 has been implicated in cancer progres- downregulated to a significantly higher degree upon THZ1 sion by activating oncogenic signalling pathways, such as RAF/ treatment, compared with those associated with TEs (figure 4F, MEK/ERK and PI3K/AKT.38 39 We were particularly interested see online supplementary table S5). These results suggested that in this candidate SE-associated kinase as its small-molecule inhi- SE-associated transcripts were particularly sensitive to partial bitors were shown to have antineoplastic activity in colon, lung, inhibition of transcription, and SE-associated oncogenes might breast and gastric cancers.40 41 However, its roles and thera- be responsible for the sensitivity of OSCC cells to low-dose peutic value in OSCC cells are yet to be explored. As mentioned THZ1 treatment. earlier, depletion of the expression of PAK4 by siRNA inhibited We hypothesised that further in-depth analysis of the expres- OSCC cell proliferation (figures 5E and 6B, see online sion dynamics of SE-associated transcripts during transcription supplementary figure S9). To examine the therapeutic value of inhibition might identify novel SE-associated oncogenes in targeting PAK4, we investigated its novel small-molecule, orally OSCC, as it is reasonable to assume that SE-associated onco- available inhibitor named KPT-9274.42 Dose–response studies genes would be strongly expressed in OSCC tumours and highly using OSCC cells expressing high PAK4 levels found most of dependent on continuous transcription. Thus, we required that them sensitive to the compound, with IC50 values ranging from the candidate novel oncogenes be: (i) associated with SEs, (ii) 180 to 612 nM (see online supplementary figure S10A). After ranked among the top 15% of all actively expressed transcripts confirming its on-target effect (figure 6D), we showed that in RNA-Seq results and (iii) highly sensitive to low-dose THZ1 KPT-9274 robustly inhibited cell proliferation and clonogenic treatment at 6 h. As shown in figure 5A, 17 candidate oncogenes growth of OSCC cells (figure 6E, see online supplementary were selected. Quantitative PCR results validated their hypersen- figure S10B). To explore the anti-OSCC effect of KPT-9274 in sitivity to transcription inhibition (figure 5B). In sharp contrast, vivo, NSG mice with KYSE510 xenografts were treated orally TE-associated genes were either only modestly decreased or with either vehicle or KPT-9274 (150 mg/kg, twice daily, 5 days/

Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 7 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus

Figure 5 Integrative analysis of chromatin immunoprecipitation sequencing (ChIP-Seq) with whole-transcriptome sequencing (RNA-Seq) results to screen for novel super-enhancer (SE)-associated oncogenes. (A) Heatmap and (B) qRT-PCR validation of RNA-Seq results of candidate SE-associated genes. (C) qRT-PCR showed that the mRNA levels of candidate genes were reduced significantly by their corresponding siRNAs (RNA from cells was harvested after 48 h of siRNA exposure). (D) H3K27ac ChIP-seq profiles of five candidate oncogenes (RUNX1, DNAJB1, YAP1, SREBF2 and PAK4) in TE7 and KYSE510 cells. (E) Left panel, western blot showing efficacy of siRNAs. Middle panel, representative results of colony formation assays of oesophageal squamous cell carcinoma (OSCC) cells transfected with indicated siRNAs. Right panel, graphic display of colony formation results. Means±SD of three independent experiments are shown (**p<0.01, ***p<0.001). (F) Immunoblotting analysis shows that the protein levels of RUNX1, YAP1, DNAJB1, SREBF2 and PAK4 were decreased by THZ1 treatment.

week). Strikingly, after 28 days, KPT-9274 almost completely DISCUSSION suppressed the tumour growth in these mice and induced Developing novel molecular-based targeted therapies is one of massive apoptosis in the xenografts (figure 6F, see online the most important strategies to treat highly aggressive cancers supplementary figures S10C–E). Importantly, no systemic tox- such as OSCC. Sadly, we and others have shown that the icity was observed during continuous administration for genomic alterations in OSCC often cause inactivation of tumour 4 weeks, and the body weight showed no significant change (see suppressor genes rather than activation of druggable onco- online supplementary figure S10F). genes.46733In this study, we demonstrated that targeting

8 Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus

Figure 6 Functional relevance of candidate super-enhancer (SE)-associated genes in oesophageal squamous cell carcinoma (OSCC) cell lines. (A) Immunoblotting analysis shows protein expression level of RUNX1, YAP1, DNAJB1, SREBF2 and PAK4 in 12 OSCC cell lines. (B) MTT (upper rows) and colony formation (lower rows) assays of indicated cells transfected with indicated siRNAs. (C) Representative immunohistochemical (IHC) results of YAP1, SREBF2 and PAK4 expression in OSCC samples and normal oesophageal epithelium. (D) OSCC cells were treated with KPT-9274 (500 nM, 48 h) and subjected to immunoblotting analysis with indicated antibodies. (E) Antiproliferation effect of KPT-9274 (500 nM) against KYSE140 and KYSE510 cells as assayed by colony formation. Representative colonies (top row) and quantification of colonies (bottom row). Data were represented as mean±SD (n=3 experiments done in triplicate, ***p<0.001). (F) OSCC (KYSE510) tumour growth curve of mice treated with either vehicle or KPT-9274. Data represent as means±SD (**p<0.01, ****p<0.0001). transcriptional regulation, as opposed to specific genomic partial transcription inhibition resulted in selective rather than lesions, might provide therapeutic value against this malignancy. global suppression of those transcripts which were highly Based on unbiased high-throughput screening with small- dependent on continuous transcription in several types of molecule inhibitors, we identified that OSCC cells displayed THZ1-sensitive tumours, but not unresponsive cancers nor non- exceptional sensitivity to CDK7 inhibition, which was exten- malignant cells.12 14 16 25 We reasoned that the exceptional sen- sively validated by in vitro and in vivo assays. We were initially sitivity of several OSCC cells to low-dose THZ1 treatment intrigued by the observations that inhibition of general gene might be conferred by a set of oncogenes which were extremely transcription by low-dose THZ1 produced much less cytotox- addicted to transcriptional activation and were thus particularly icity in healthy tissues than in OSCC xenograft (see online vulnerable to CDK7 inhibition. Indeed, RNA-seq profiles supplementary figure S5). Recent investigations found that showed that low-dose THZ1 treatment led to selective

Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 9 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus inhibition of a number of well-known OSCC oncogenes, many Qi et al53 found that it could decrease cell proliferation in a of which are transcriptional factors, such as TP63,31 CTTN,36 p53-dependent manner in breast cancers. Our data revealed that SOX2,43 PLK1,44 STAT345 and ID146 (see online as an SE-associated oncogene, DNAJB1 was highly expressed in supplementary tables S2 and S3). Interestingly, we observed that OSCC compared with other human cancers (see online many non-coding RNAs with high-level expression were also supplementary figure S11), and it significantly promoted the particularly vulnerable to THZ1 treatment, such as growth and proliferation of OSCC cells. MIR205HG, TP53TG1 and MIR210HG, whose functions are Last, our systematic approach identified a druggable still unknown. Therefore, our approach may help nominate SE-associated oncogene, PAK4. Both in vitro and in vivo experi- novel oncogenic non-coding RNAs in OSCC cells. ments confirmed that its small-molecule inhibitor, KPT-9274, Although a set of oncogenes with extreme addiction to RNA dramatically suppressed OSCC cell viability and induced Pol-II mediated transcription might explain the sensitivity of massive apoptosis. These data suggested the potential thera- OSCC cells to CDK7 inhibition, we still wondered what caused peutic value of targeting PAK4 for clinical management of the continuous transcriptional activation of these oncogenes in patients with OSCC. OSCC cells. To probe the mechanism, we first performed GO In aggregate, the current study addressed both basic and analysis and found that THZ1-sensitive transcripts were translational questions, which are all highly novel and unex- enriched in processes that regulate gene transcription, DNA plored in the context of OSCC biology. Specifically, our results repair and cell apoptosis, all of which are critical for cancer provide an important molecular foundation to understand the biology. As oncogenes involved in these cellular functions have transcriptional landscape of OSCC and a catalogue of novel been shown to be associated with SEs in other types of oncogenic transcripts, both of which are valuable for the OSCC tumours, we hypothesised that SEs might be responsible for the research community. Moreover, our work may help establish the robust transcriptional activation of THZ1-sensitive transcripts in therapeutic merit of targeting SE-associated oncogenic transcrip- OSCC cells. tion programme for OSCC treatment. Hnisz et al10 have created a catalogue of SEs for different types of human cells and tissues, OSCC was not investigated. Author affiliations 1 Here, we characterised the SE landscape of OSCC, and identi- Cancer Science Institute of Singapore, National University of Singapore, Singapore, fi Singapore ed that SE-associated transcripts are enriched in processes 2Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene which are exclusive to SCC biology, such as epidermis develop- Regulation, Medical Research Center, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen ment and keratinocyte differentiation, as well as cell prolifer- University, Guangzhou, China 3 ation and apoptosis. Interestingly, we also discovered some Department of Drug Discovery, Karyopharm Therapeutics Inc., Newton, novel SE-associated non-coding mRNAs. For example, Massachusetts, USA 4State Key Laboratory of Molecular Oncology, Cancer Institute (Hospital), Peking MIR205HG was associated with one of the largest SEs in Union Medical College and Chinese Academy of Medical Sciences, Beijing, China OSCC cells. Expression profiling also supported the uniqueness 5Department of Cell, Developmental & Cancer Biology, Knight Cancer Institute, of this non-coding RNA in SCC. Oregon Health & Science University, Portland, Oregon, USA 6 We reasoned that as both THZ1-sensitive, SE-associated Division of Hematology/Oncology, Cedars-Sinai Medical Center, University of California, Los Angeles School of Medicine, Los Angeles, California, USA cohort of transcripts contain oncogenic factors important for 7National University Cancer Institute, National University Health System and National OSCC malignant phenotype, an integrative interrogation of University of Singapore, Singapore, Singapore these two datasets might identify novel candidate oncogenes. Through a series of functional experiments, we showed that 5 Acknowledgements The authors thank Anand D Jeyasekharan (a member of out 17 candidate SE-associated genes, namely, RUNX1, YAP1, Singapore Gastric Cancer Consortium) for his helpful advice. The authors also thank Hazimah Binte Mohd Nordin for help with mouse work. DNAJB1, PAK4 and SREBF2, were important for the prolifer- ation of OSCC cells. In addition, we confirmed the overexpres- Contributors D-CL, Y-YJ and HPK designed the study, analysed and discussed the data and wrote the manuscript. Y-YJ, MH, LX, L-WD performed experiments. LS, XX sion of some of these proteins in OSCC compared with and M-RW coordinated sample collection and processing. WS and EB provided non-malignant oesophageal epithelium. Since we only tested the KPT-9274 and discussed its related results. AM and HY performed bioinformatical phenotypes of cell proliferation, candidates implicated in other analysis. JWT helped conduct high-throughput small-molecule inhibitor screening. cellular functions such as migration, invasion and epithelial– YC, J-FX, DK and W-WC discussed the data. mesenchymal transition might be overlooked and need further Funding This work was funded by the Singapore Ministry of Health’s National investigations to characterise. Medical Research Council (NMRC) under its Singapore Translational Research (STaR) Investigator Award to HPK and by NMRC Individual Research Grant (NMRC/1311/ Transcription factor RUNX1 is a well-studied differentiation 2011) and National Natural Science Foundation of China (81520108023, regulator and tumour suppressor in haematopoietic cells. 81330052) to M-RW. D-CL was supported by Donna and Jesse Garber Awards for However, its functions in normal and malignant epithelial cells Cancer Research and National Center for Advancing Translational Sciences UCLA remain obscure and inconclusive. For example, it was reported CTSI Grant UL1TR000124. This study was partially supported by a generous to promote the proliferation of skin squamous cancer cells47 48 donation from the Melamed family. but inhibit breast cancer cells.49 Interestingly, RUNX1 gene is Competing interests WS and EB are employees of Karyopharm Therapeutics Inc. frequently deleted in EA and RUNX1 suppressed the prolifer- Provenance and peer review Not commissioned; externally peer reviewed. ation of EA cells.50 51 In sharp contrast, here we show that RUNX1 is an SE-associated oncogene and promotes cell prolif- eration in OSCC. These results again underscore the ability of REFERENCES fi 1 He YT, Hou J, Chen ZF, et al. [Study on the esophageal cancer incidence and our integrative approaches to discern cell type-speci c gene mortality rate from 1974–2002 in Cixian, China]. Zhonghua liu xing bing xue za functions. zhi 2006;27:127–31. Similarly, DNAJB1 is poorly studied in human cancers and 2 Zhao P, Dai M, Chen W, et al. Cancer trends in China. Jpn J Clin Oncol appears to have seemingly opposite roles. Specifically, as a 2010;40:281–5. 3 Pennathur A, Gibson MK, Jobe BA, et al. Oesophageal carcinoma. Lancet protein implicated in stimulating the ATPase activity of Hsp70s, 2013;381:400–12. investigators showed that DNAJB1 inhibited p53-mediated 4 Lin DC, Hao JJ, Nagata Y, et al. Genomic and molecular characterization of apoptosis by destabilising PDCD5 in lung cancer.52 In contrast, esophageal squamous cell carcinoma. Nat Genet 2014;46:467–73.

10 Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com Oesophagus

5 Lin DC, Wang MR, Koeffler HP. Targeting genetic lesions in esophageal cancer. Cell 31 Watanabe H, Ma Q, Peng S, et al. SOX2 and p63 colocalize at genetic loci in Cycle 2014;13:2013–14. squamous cell carcinomas. J Clin Invest 2014;124:1636–45. 6 Gao YB, Chen ZL, Li JG, et al. Genetic landscape of esophageal squamous cell 32 Liu R, Gu J, Jiang P, et al. DNMT1-microRNA126 epigenetic circuit contributes to carcinoma. Nat Genet 2014;46:1097–102. esophageal squamous cell carcinoma growth via ADAM9-EGFR-AKT signaling. Clin 7 Song Y, Li L, Ou Y, et al. Identification of genomic alterations in oesophageal Cancer Res 2015;21:854–63. squamous cell cancer. Nature 2014;509:91–5. 33 Shi ZZ, Shang L, Jiang YY, et al. Consistent and differential genetic aberrations 8 Agrawal N, Jiao Y, Bettegowda C, et al. Comparative genomic analysis of between esophageal dysplasia and squamous cell carcinoma detected by array esophageal adenocarcinoma and squamous cell carcinoma. Cancer Discov comparative genomic hybridization. Clin Cancer Res 2013;19:5867–78. 2012;2:899–905. 34 Gen Y, Yasui K, Nishikawa T, et al. SOX2 promotes tumor growth of esophageal 9 Lovén J, Hoke HA, Lin CY, et al. Selective inhibition of tumor oncogenes by squamous cell carcinoma through the AKT/mammalian target of rapamycin complex disruption of super-enhancers. Cell 2013;153:320–34. 1 signaling pathway. Cancer Sci 2013;104:810–16. 10 Hnisz D, Abraham BJ, Lee TI, et al. Super-enhancers in the control of cell identity 35 Xie JJ, Xu LY, Zhang HH, et al. Role of fascin in the proliferation and invasiveness and disease. Cell 2013;155:934–47. of esophageal carcinoma cells. Biochem Biophys Res Commun 2005; 11 Chapuy B, McKeown MR, Lin CY, et al. Discovery and characterization of 337:355–62. super-enhancer-associated dependencies in diffuse large B cell lymphoma. Cancer 36 Luo ML, Shen XM, Zhang Y, et al. Amplification and overexpression of CTTN cell 2013;24:777–90. (EMS1) contribute to the metastasis of esophageal squamous cell carcinoma by 12 Christensen CL, Kwiatkowski N, Abraham BJ, et al. Targeting transcriptional promoting cell migration and anoikis resistance. Cancer Res 2006;66: addictions in small cell lung cancer with a covalent CDK7 inhibitor. Cancer cell 11690–9. 2014;26:909–22. 37 Song JH, Meltzer SJ. Long non-coding RNA SS-001 regulates proliferation, cell cycle, 13 Amaral PP, Bannister AJ. Re-place your BETs: the dynamics of super enhancers. Mol and migration in esophageal adenocarcinoma cells. Cancer Res 2013;73(8 Cell 2014;56:187–9. Suppl):1839. 14 Chipumuro E, Marco E, Christensen CL, et al. CDK7 inhibition suppresses 38 Siu MK, Chan HY, Kong DS, et al. p21-activated kinase 4 regulates ovarian cancer super-enhancer-linked oncogenic transcription in MYCN-driven cancer. Cell cell proliferation, migration, and invasion and contributes to poor prognosis in 2014;159:1126–39. patients. Proc Natl Acad Sci USA 2010;107:18622–7. 15 Mansour MR, Abraham BJ, Anders L, et al. Oncogene regulation. An oncogenic 39 Tabusa H, Brooks T, Massey AJ. Knockdown of PAK4 or PAK1 inhibits the super-enhancer formed through somatic mutation of a noncoding intergenic proliferation of mutant KRAS colon cancer cells independently of RAF/MEK/ERK and element. Science 2014;346:1373–7. PI3K/AKT signaling. Mol Cancer Res 2013;11:109–21. 16 Kwiatkowski N, Zhang T, Rahl PB, et al. Targeting transcription regulation in cancer 40 Murray BW, Guo C, Piraino J, et al. Small-molecule p21-activated kinase inhibitor with a covalent CDK7 inhibitor. Nature 2014;511:616–20. PF-3758309 is a potent inhibitor of oncogenic signaling and tumor growth. Proc 17 Tyner JW, Yang WF, Bankhead A III, et al. Kinase pathway dependence in primary Natl Acad Sci USA 2010;107:9446–51. human leukemias determined by rapid inhibitor screening. Cancer Res 41 Zhang J, Wang J, Guo Q, et al. LCH-7749944, a novel and potent p21-activated 2013;73:285–96. kinase 4 inhibitor, suppresses proliferation and invasion in human gastric cancer 18 Bicocca VT, Chang BH, Masouleh BK, et al. Crosstalk between ROR1 and the Pre-B cells. Cancer Lett 2012;317:24–32. cell receptor promotes survival of t(1;19) acute lymphoblastic leukemia. Cancer Cell 42 Senapedis W, Landesman Y, Schenone M, et al. Identification of Novel Small 2012;22:656–67. Molecules as Selective PAK4 Allosteric Modulators (PAMs) by Stable Isotope 19 Huang Y, Xi Q, Chen Y, et al. A dual mTORC1 and mTORC2 inhibitor shows Labeling of Amino acids in Cells (SILAC). Eur J Cancer 2014;50:156. antitumor activity in esophageal squamous cell carcinoma cells and sensitizes them 43 Bass AJ, Watanabe H, Mermel CH, et al. SOX2 is an amplified lineage-survival to cisplatin. Anticancer Drugs 2013;24:889–98. oncogene in lung and esophageal squamous cell carcinomas. Nat Genet 20 Liu SG, Wang BS, Jiang YY, et al. Atypical protein kinase Ciota (PKCiota) promotes 2009;41:1238–42. metastasis of esophageal squamous cell carcinoma by enhancing resistance to 44 Zhang Y, Du XL, Wang CJ, et al. Reciprocal activation between PLK1 and Stat3 Anoikis via PKCiota-SKP2-AKT pathway. Mol Cancer Res 2011;9:390–402. contributes to survival and proliferation of esophageal cancer cells. Gastroenterology 21 Wang XT, Bao CH, Jia YB, et al. BIIB021, a novel Hsp90 inhibitor, sensitizes 2012;142:521–30 e3. esophageal squamous cell carcinoma to radiation. Biochem Biophys Res Commun 45 Du XL, Yang H, Liu SG, et al. Calreticulin promotes cell motility and enhances 2014;452:945–50. resistance to anoikis through STAT3-CTTN-Akt pathway in esophageal squamous 22 Ui T, Morishima K, Saito S, et al. The HSP90 inhibitor 17-N-allylamino-17- cell carcinoma. Oncogene 2009;28:3714–22. demethoxy geldanamycin (17-AAG) synergizes with cisplatin and induces apoptosis 46 Li B, Tsao SW, Li YY, et al. Id-1 promotes tumorigenicity and metastasis of human in cisplatin-resistant esophageal squamous cell carcinoma cell lines via the Akt/XIAP esophageal cancer cells through activation of PI3K/AKT signaling pathway. Int J pathway. Oncol Rep 2014;31:619–24. Cancer 2009;125:2576–85. 23 Sato F, Kubota Y, Natsuizaka M, et al. EGFR inhibitors prevent induction of cancer 47 Scheitz CJ, Lee TS, McDermitt DJ, et al.Defining a tissue stem cell-driven Runx1/ stem-like cells in esophageal squamous cell carcinoma by suppressing Stat3 signalling axis in epithelial cancer. EMBO J 2012;31:4124–39. epithelial-mesenchymal transition. Cancer Biol Ther 2015;16:933–40. 48 Hoi CS, Lee SE, Lu SY, et al. Runx1 directly promotes proliferation of hair follicle 24 Chen J, Lan T, Zhang W, et al. Dasatinib enhances cisplatin sensitivity in human stem cells and epithelial tumor formation in mouse skin. Mol Cell Biol esophageal squamous cell carcinoma (OSCC) cells via suppression of PI3K/AKT and 2010;30:2518–36. Stat3 pathways. Arch Biochem Biophys 2015;575:38–45. 49 Wang L, Brugge JS, Janes KA. Intersection of FOXO- and RUNX1-mediated gene 25 Wang Y, Zhang T, Kwiatkowski N, et al. CDK7-Dependent Transcriptional Addiction expression programs in single breast epithelial cells during morphogenesis and in Triple-Negative Breast Cancer. Cell 2015;163:174–86. tumor progression. Proc Natl Acad Sci USA 2011;108:E803–12. 26 Larochelle S, Amat R, Glover-Cutter K, et al. Cyclin-dependent kinase control of the 50 Gu J, Ajani JA, Hawk ET, et al. Genome-wide catalogue of chromosomal initiation-to-elongation switch of RNA polymerase II. Nat Struct Mol Biol aberrations in barrett’s esophagus and esophageal adenocarcinoma: a high-density 2012;19:1108–15. single nucleotide polymorphism array analysis. Cancer Prev Res (Phila) 27 Zhou Q, Li T, Price DH. RNA polymerase II elongation control. Annu Rev Biochem 2010;3:1176–86. 2012;81:119–43. 51 Dulak AM, Schumacher SE, van Lieshout J, et al. Gastrointestinal adenocarcinomas 28 Oishi Y, Manabe I, Tobe K, et al. Krüppel-like transcription factor KLF5 is a key of the esophagus, stomach, and colon exhibit distinct patterns of genome instability regulator of adipocyte differentiation. Cell Metab 2005;1:27–39. and oncogenesis. Cancer Res 2012;72:4383–93. 29 Truong AB, Kretz M, Ridky TW, et al. p63 regulates proliferation and differentiation 52 Cui X, Choi HK, Choi YS, et al. DNAJB1 destabilizes PDCD5 to suppress of developmentally mature keratinocytes. Genes Dev 2006;20:3185–97. p53-mediated apoptosis. Cancer Lett 2015;357:307–15. 30 Richardson RJ, Dixon J, Malhotra S, et al. Irf6 is a key determinant of the 53 Qi M, Zhang J, Zeng W, et al. DNAJB1 stabilizes MDM2 and contributes to cancer cell keratinocyte proliferation-differentiation switch. Nat Genet 2006;38:1329–34. proliferation in a p53-dependent manner. Biochim Biophys Acta 2014;1839:62–9.

Jiang Y-Y, et al. Gut 2016;0:1–11. doi:10.1136/gutjnl-2016-311818 11 Downloaded from http://gut.bmj.com/ on May 23, 2016 - Published by group.bmj.com

Targeting super-enhancer-associated oncogenes in oesophageal squamous cell carcinoma Yan-Yi Jiang, De-Chen Lin, Anand Mayakonda, Masaharu Hazawa, Ling-Wen Ding, Wen-Wen Chien, Liang Xu, Ye Chen, Jin-Fen Xiao, William Senapedis, Erkan Baloglu, Deepika Kanojia, Li Shang, Xin Xu, Henry Yang, Jeffrey W Tyner, Ming-Rong Wang and H Phillip Koeffler

Gut published online May 10, 2016

Updated information and services can be found at: http://gut.bmj.com/content/early/2016/05/10/gutjnl-2016-311818

These include: References This article cites 53 articles, 16 of which you can access for free at: http://gut.bmj.com/content/early/2016/05/10/gutjnl-2016-311818#BIBL

Email alerting Receive free email alerts when new articles cite this article. Sign up in the service box at the top right corner of the online article.

Topic Articles on similar topics can be found in the following collections Collections Oesophageal cancer (349)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/